What is Dexter?
Dexter is an open-source autonomous agent for deep financial research, designed to turn complex financial questions into structured investigations. Dexter combines task planning, live market data, and self-reflection to gather evidence and refine answers. The project is available through a public GitHub repository and is licensed under the MIT License.
Dexter is intended for educational, entertainment, and informational purposes rather than real trading or investment. The documentation states that Dexter does not provide financial, investment, tax, or legal advice. Dexter's outputs may be incorrect, incomplete, or out of date, so the research workflow should not be treated as a guarantee of accuracy.
What Dexter can do
Dexter organizes financial research around a sequence of planned tasks rather than a single response. Dexter selects tools to gather financial data, checks the work, and iterates on results. The documented financial data access includes income statements, balance sheets, and cash flow statements, while execution safeguards address repeated loops and excessive steps.
- Dexter decomposes complex financial queries into structured research steps.
- Dexter selects and executes tools to gather financial data.
- Dexter checks its own work and iterates until tasks are complete.
- Dexter accesses income statements, balance sheets, and cash flow statements.
- Dexter uses loop detection and step limits to constrain execution.
- Dexter logs tool calls and reasoning steps in query-specific scratchpad files.
How Dexter works
Dexter runs after the repository is cloned, dependencies are installed with Bun, and environment variables are configured in a local .env file. The prerequisites list OpenAI and Financial Datasets API keys, with Exa optional for web search. The configuration examples also include other model providers, a local Ollama endpoint, and Tavily as a search fallback.
Dexter starts in interactive mode with bun start, while bun dev enables watch mode for development. Dexter also supports WhatsApp through a gateway: users link their account by scanning a QR code, start the gateway, and send questions to their own chat. Dexter returns responses to that same conversation.
Who Dexter is for
Dexter fits learners exploring financial research workflows and developers examining how an autonomous research agent plans, retrieves data, and validates results. Dexter includes debugging records and an evaluation suite, making the repository relevant to people testing agent behavior as well as asking financial questions. These uses remain subject to the project's educational and informational restrictions.
- Dexter can support educational investigations of company financial statements.
- Dexter can break complex financial questions into research tasks.
- Dexter can answer research questions submitted through WhatsApp.
- Dexter can be evaluated against a dataset of financial questions.
- Dexter lets developers inspect retrieved data, tool arguments, and generated summaries.
Dexter pricing and access
Dexter is an open-source project distributed through GitHub under the MIT License. The repository documentation describes installation and configuration rather than subscription plans or hosted product tiers. Dexter requires a Bun runtime and configured access to financial data and model services; the source does not specify prices for those external services.
Dexter strengths and limitations
Dexter combines autonomous research execution with records that developers can inspect. Each query produces a JSONL scratchpad containing the original question, tool results, and reasoning steps. Dexter also includes an evaluation suite that tracks results in LangSmith and uses an LLM judge to score correctness, providing a documented workflow for examining agent performance.
- Dexter requires local installation and environment configuration before use.
- Dexter depends on configured model and financial data access.
- Dexter is not intended for real trading or investment.
- Dexter does not guarantee accurate, complete, or current outputs.
Dexter FAQ
What is Dexter?
Dexter is an open-source autonomous financial research agent that turns complex questions into research plans, gathers live market data, and checks its work. Dexter is intended for educational, entertainment, and informational purposes, not real trading or investment.
Is Dexter free?
Dexter is available as an open-source project under the MIT License, rather than a documented subscription product. Dexter's setup requires configured model and financial data access, but the repository text does not state prices for external API services.
How do you install and run Dexter?
Dexter runs locally using Bun after users clone the GitHub repository, install dependencies, and configure environment variables. Dexter starts in interactive mode with bun start. The prerequisites list OpenAI and Financial Datasets API keys, with Exa optional for web search.
Can you use Dexter through WhatsApp?
Dexter supports WhatsApp through a gateway linked to the user's account. Dexter's setup involves scanning a QR code with the gateway login command and starting the gateway. Questions sent to the user's own WhatsApp chat are processed, and responses return to that conversation.
What financial data can Dexter access?
Dexter accesses live financial data, including income statements, balance sheets, and cash flow statements. Dexter selects tools to gather information for a research question, then checks and refines the results as part of its autonomous research workflow.
Can Dexter be used for investment advice?
Dexter is not intended for real trading or investment and does not provide financial, investment, tax, or legal advice. Dexter's documentation warns that outputs may be incorrect, incomplete, or out of date and advises consulting a licensed financial advisor before investment decisions.
From the community
No reviews yet
Be the first to review this tool. What you write helps the next person choose.




